The current implementation using a Nema Size 2 Contactor and a Zero Cross SSR causes a spike activation on the coil when switching from OFF to AUTO. Specifications on the SSR state that a 10mA leak occurs but the activation is very temporary ( mS ); not even enough to turn the motor but the motor does jolt just for a second. The symptom isn't even accurately repeatable. Sometimes no jolt happens.
I'd like to see if the Random type SSR ( recommended for inductive loads ) will solve this symptom - however the system does seem to work just not as ideally as expected.